Evaluation method of user comprehensive influence based on analytic hierarchy process

Weibo users with high-impact play an important role in promoting information dissemination, network marketing, and even guiding the trend of public opinion, which is of great significance to the study of social applications. Because of concerning the timelines shortage and the problems of incomplete...

Full description

Saved in:
Bibliographic Details
Published in2021 International Conference on Artificial Intelligence and Electromechanical Automation (AIEA) pp. 348 - 355
Main Authors Sen, Wang, Zhang, Bofeng, Li, Tingting, Lv, Hehe, Hu, Shengxiang, Peng, Yubo
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.05.2021
Subjects
Analytic hierarchy process
Authentication
Automation
blog characteristics
Blogs
Fans
half-life
Heuristic algorithms
user characteristics
user influence
Online AccessGet full text

Cover

More Information
Summary:Weibo users with high-impact play an important role in promoting information dissemination, network marketing, and even guiding the trend of public opinion, which is of great significance to the study of social applications. Because of concerning the timelines shortage and the problems of incomplete analysis of blog's factors in measuring user influence, a user influence method, named User Comprehensive Influence Rank(UCIR), was proposed. The method defines the static influence of user according to user's characteristics such as the number of fans, Weibo authentication, network centrality, etc, and the dynamic influence of blog based on blog's characteristics such as the number of forwards, comments, praises and time factors. Firstly, it calculates the weight of user characteristics and blog characteristics through analytic hierarchy process and user static influence is calculated by adding the weight of each feature. Then blog dynamic influence is calculated by introducing the concept of half-life. Finally, the comprehensive influence of Weibo user is calculated by considering static influence of user and dynamic influence of blog. Compared with TwitterRank algorithms and PageRank algorithms, UCIR algorithm improves the precision and recall by 28.7%, 50.9% and 32.5%, 60.2% respectively, which proves the effectiveness of UCIR algorithm. This method can more accurately evaluate the real user influence under a specific topic.
DOI:10.1109/AIEA53260.2021.00080